To establish the usefulness of liver slices as predictive models for in vivo metabolism and toxicity, acetaminophen was used as a model hepatotoxicant for which the role of metabolism in toxicity is well documented. Acetaminophen was incubated with liver slices prepared from the rat and hamster since these species differ in susceptibility to acetaminophen-induced hepatotoxicity. Formation of acetaminophen metabolites (sulfate, glucuronide, and glutathione conjugate), slice glutathione levels, and slice histopathology were assessed. Acetaminophen (0.5, 1, and 2 mM) induced a dose-dependent depletion of glutathione in hamster slices (sensitive species) but not rat slices (insensitive species). Formation of the acetaminophen toxic metabolite, as measured by glutathione conjugate formation, was much lower in rat slices compared with hamster slices. Rat slices also showed greater activity of the nontoxic sulfation and glucuronidation pathways. These data predicted an 11-fold greater susceptibility to toxicity in the hamster based on the proportion of the dose metabolized to the reactive species. In vivo data, using hepatic glutathione depletion as an indicator of toxic metabolite formation, would predict a similar difference (13.3-fold) between the hamster and the rat. To ascertain if the characteristic centrilobular lesion induced by acetaminophen in vivo could be reproduced in vitro, liver slice histopathology was assessed 6 and 12 hr after a 2-hr treatment with acetaminophen (2 mM). Discrete damage to the centrilobular regions of the liver were noted in hamster but not rat slices. Overall, based on metabolite formation, glutathione depletion, and slice histopathology, the liver slices were excellent predictors of acetaminophen hepatotoxicity in vivo.